Diets in low-SES Hispanic children were adequate in most nutrients, but frequently exceeded guidelines for percent total fat, percent saturated fat, cholesterol, added sugar, and sodium intakes. Low intakes of fruits, vegetables, and fiber diminished the quality of the diet. With few exceptions (noted in the results section), non-overweight and overweight children from these Hispanic families consumed diets that were qualitatively, but not quantitatively, similar. This finding may result from the availability of similar foods within a shared household.
In the present VIVA study, diet quality of both the non-overweight and overweight children did not adhere to the US Dietary Guidelines which recommend the consumption of a variety of nutrient-dense foods and beverages within and among the basic food groups while choosing foods that limit the intake of saturated and trans fats, cholesterol, added sugars and salt (9
). The principal food sources providing 68% of dietary energy included soda, desserts, pizza, snack chips, fruit drinks, fruit juice, processed meats and burgers which are high in fat, sugar, and/or sodium. The high reliance on grains, cold cereal, fruit juice, and fruit drinks suggests that the nutrient adequacy of the diets of these children was augmented by food fortification. Interestingly, the percent of energy from cold cereal and whole milk consumed by overweight children was less than that consumed by non-overweight children. Overweight children tended to consume greater amounts of low-fat milks, so that only the percent of energy varied, consistent with similar numbers of milk servings between groups. The association between dairy intake and weight status is not clear in the literature (30
). Conflicting results in the literature may be attributed to differences in study design and subject profiles, use of all dairy food products versus milk, and use of milks varying in fat content. Further exploration of the association between dairy product consumption and weight status is needed to better interpret the results from the current study.
The mean number of servings of fruits and vegetables was lower than recommended in the US Dietary Guidelines for all children. A dietary pattern consisting of low intakes of fruits and vegetables, together with high intakes of high-fat, high-sugar foods, is consistent with that found in previous studies in Mexican-American children (31
). However, the relatively higher vegetable intake among overweight children, compared with their non-overweight siblings, maybe due to eating greater amounts of raw and cooked vegetables or combination foods prepared with vegetables.
The relatively low proportion of restaurant/cafeteria/fast foods consumed by the VIVA cohort, as reported in the Results section, is in contrast to findings from the Continuing Survey of Food Intakes by Individuals (CSFII) 1994-1996 in which 30.3% of all children and 23.3% of Hispanic children reported consuming fast food on a typical day (33
). Limited income or cultural practices may have limited the number of meals outside of the home for the VIVA families.
In other aspects, the food consumption patterns of the VIVA children were not qualitatively dissimilar from other American children. The macronutrient composition of the diets of these Hispanic children resembles that of the general US population of children. Data collected from NHANES 2003-2004 (34
) were not different for Mexican-Americans versus all ethnic/racial groups (6-19 y): 53%, 14%, and 33% of energy from carbohydrate, protein, and fat, respectively. These percentages are similar to those attained in this study, although total energy intake in VIVA was lower by 200 kcals. Qualitative results similar to those in the current study were found in yet another study. The macronutrient distribution, 4-18 y old, was 14% of energy from protein, 54% from carbohydrate, and 32% from fat in the Continuing Survey of Food Intakes by Individuals (CSFII) 1994-1996, 1998 (29
). The mean cholesterol intake was 232 g and the mean fiber intake was 13.7 g. Percent of energy from added sugar was not comparable between the VIVA children and those in the CSFII sample, however, where intakes were 20.3% and 15.5%, respectively (35
). The reason for this is not clear. Differences between VIVA and NHANES or CSFII data sets could result from regional influences unique to southeast Texas, socio-economic parameters, or acculturation into Hispanic traits and social patterns. Alternatively, an increase of added sugar intake in the overall population in more recent years could account for the difference. Milk consumption decreases with age, coincident with the rise in soda intake (36
In the VIVA study, the percent of energy intake from sodas (6.3%) and fruit drinks (3.7%) did not differ between the non-overweight and overweight children. According to the Nationwide Food Consumption Survey 1977/1978, CSFII 1994/1996, and the Supplemental Children's Survey 1998
, the mean intake of soda in children, 6-17 y, increased from 5 to 12 fl oz per day in 1977/78 to 1994/98; energy derived from soda doubled from 2.9 to 5.9% of total energy intake (38
Correlations between percent of energy from food groups and nutrient z-scores were consistent with expected food composition values, with few exceptions. Iron z-scores were not significantly correlated with meat, but were positively correlated with percent of energy from grains, which likely reflects fortification of cereals consumed. Further evidence of dietary fortification was the positive correlation between percent of energy from grains and z-scores for vitamin E, thiamin, riboflavin, niacin, and folate and also between percent of energy from dairy and z-scores of vitamin A and vitamin D. The lack of an association between percent of energy from meat and protein z-score was unexpected; however, protein z-score was calculated per kilogram body weight which may explain why the anticipated result was not observed. Most notably, in these Hispanic children, intake of energy in the form of beverages was negatively associated with z-scores for a majority of nutrients; only z-scores for vitamin C and sodium were not associated. Nutrient dilution results when energy-dense and nutrient-poor foods, frequently sugar-laden beverages, are consumed to the extent that the nutritive quality of the diet is reduced. The beverages food group did not contain milk or fruit juice, but was comprised chiefly of drinks containing added sugars. The negative association between energy from beverages and nutrient z-scores corroborates findings in other studies where increasing intake of sweetened beverages was associated with nutrient dilution (35
). However, the positive correlations between percent of energy from dairy and many nutrient z-scores substantiates the contribution of dairy products to nutrient adequacy.
The nutrient intakes of these Hispanic children were adequate, for the most part. Exceptions included vitamin E which was low relative to the EAR and vitamin D, pantothenic acid, calcium and potassium which were lower than the AI. Fiber intake also was considerably below the AI. These are nutrients or dietary components that are commonly reported low in US diets (41
). Definitive assessment of the nutrient adequacy or inadequacy of vitamin D, pantothenic acid, calcium and potassium is not possible given the uncertainty of the AI. For the determination of the EAR for vitamin E and the AI for fiber, there were no substantive pediatric data to determine requirements and therefore the extrapolation of adult data to children may be inappropriate.
This study has some limitations. First, the results of this study are only generalizable to low-SES Hispanic children living in an urban community. Secondly, since these data were cross-sectional, cause and effect relationships cannot be determined. Longitudinal studies would be needed to test the effect of diet on the development of obesity.